CN117794447A

CN117794447A - Method for targeting a personalized neuromodulated neural circuit for impulse-related or uncontrollable behavior

Info

Publication number: CN117794447A
Application number: CN202280054546.8A
Authority: CN
Inventors: D·A·N·巴尔博萨; C·哈尔彭; J·麦克纳布
Original assignee: Leland Stanford Junior University
Current assignee: Leland Stanford Junior University
Priority date: 2021-06-14
Filing date: 2022-06-13
Publication date: 2024-03-29
Also published as: WO2022266000A1

Abstract

Provided herein is a method of targeting a neural circuit of a subject having impulse-related or uncontrollable behavior, the method comprising (a) generating a circuit-specific brain connectivity spectrum using a pattern of water diffusivity based on an MRI of the subject; (b) Generating a loop-specific voxel-by-voxel connection matrix with the loop-specific brain connection spectrum to define a region of interest including a neural loop to be targeted; (c) Segmenting the loop-specific voxel-by-voxel connection matrix using an algorithm to generate 3D subregions to identify neural loops to be targeted; and (d) targeting the identified neural circuit with neuromodulation therapy.

Description

Method for targeting a personalized neuromodulated neural circuit for impulse-related or uncontrollable behavior

Cross Reference to Related Applications

The present application claims priority from U.S. provisional patent application Ser. No. 63/210,472 filed on 6/14 of 2021, which is incorporated herein by reference in its entirety for all purposes.

Background

Obesity and binge eating, a behavior characterized by pathological impulses and uncontrollable behavior, coexist as strong predictors of treatment refractory, even the most aggressive treatments result in the worst outcome (McCuen-Wurst et al, 2018; white et al, 2010). Elucidating the neural relevance of this common complication may provide unique insight into the mechanism of treatment difficulty and may guide the development of neuromodulation therapies.

The ventral inner prefrontal cortex (vmPFC) to nucleus accumbens (NAc) circuit is involved in inhibitory control. Impaired inhibitory control associated with obesity is often manifested as binge eating, which is associated with a higher incidence of metabolic and psychiatric disorders and poor therapeutic outcome. It is not clear whether there is an abnormality in this inhibitory control loop in subjects with a binge eating trend that are obese.

The present disclosure may not only elucidate the neurological relevance of binge eating and obesity complications, but may also inform and address the urgent need to develop circuit-based therapies for this major unmet need.

Disclosure of Invention

Provided herein is a method of targeting a neural circuit of a subject having impulse-related or uncontrollable behavior, the method comprising (a) generating a circuit-specific brain connection profile using a pattern of water diffusivity based on an MRI of the subject; (b) Generating a loop-specific voxel-by-voxel connection matrix with the loop-specific brain connection spectrum to define a region of interest including a neural loop to be targeted; (c) Segmenting the loop-specific voxel-by-voxel connection matrix using an algorithm to generate 3D subregions to identify neural loops to be targeted; and (d) targeting the identified neural circuit with neuromodulation therapy.

Provided herein is a method of treating a subject having impulse-related or uncontrollable behavior, the method comprising (a) generating a circuit-specific brain connection profile using a pattern of water diffusivity based on an MRI of the subject; (b) Generating a loop-specific voxel-by-voxel connection matrix with the loop-specific brain connection spectrum to define a region of interest including a neural loop to be targeted; (c) Segmenting the loop-specific voxel-by-voxel connection matrix using an algorithm to generate 3D subregions to identify neural loops to be targeted; and (d) treating the subject with neuromodulation therapy by targeting the identified neural circuit.

Provided herein is a method of personalized neuromodulation therapy for a subject having impulse-related or uncontrollable behavior, the method comprising (a) generating a circuit specific brain connection profile using a pattern of water diffusivity based on an MRI of the subject; (b) Generating a loop-specific voxel-by-voxel connection matrix with the loop-specific brain connection spectrum to define a region of interest including a neural loop to be targeted; (c) Segmenting the loop-specific voxel-by-voxel connection matrix using an algorithm to generate 3D subregions to identify neural loops to be targeted; and (d) personalizing the neuromodulation therapy based on patient-specific neurommaging, brain connectivity, cortical thickness, clinical and behavioral complaints, previous responses to neuromodulation, and other available clinical and behavioral covariates of the subject.

In some embodiments, the MRI is diffusion, structural, and/or functional resting state MRI.

In some embodiments, the algorithm is selected from the group consisting of k-means clustering algorithms, independent component analysis, principal component analysis, random relaxation with decoder perturbation, inter-centroid random search, and hierarchical clustering algorithms.

In some embodiments, the region of interest is selected from the group consisting of nucleus accumbens (NAc), inner capsular forelimbs (ALIC), subthalamic nucleus (STN), thalamus, ventral medial prefrontal cortex (vmPFC), prefrontal cinal cortex (ACC), orbitofrontal (OFC)/frontal polar cortex (FP), and dorsal medial/dorsal lateral prefrontal cortex. In some embodiments, the region of interest is a nucleus accumbens (NAc).

In some embodiments, the identified neural circuit is selected from the group consisting of an ventral prefrontal cortex (vmPFC) -NAc circuit, an orbital/frontal cortex-ALIC circuit, an orbital/frontal cortex-thalamus circuit, a vmPFC-NAc circuit, an ACC-NAc circuit, a dorsal prefrontal cortex-vmPFC-NAc circuit. In some embodiments, the identified neural circuit is a vmPFC-NAc circuit. In some embodiments, the neuromodulation therapy is invasive or non-invasive.

In some embodiments, the neuromodulation therapy is selected from deep brain stimulation, ultrasound focusing, transcranial magnetic stimulation, transcranial electrical stimulation, radio modulation, and nerve ablation.

In some embodiments, the impulse-related or uncontrollable behavior is selected from the group consisting of (1) pathological impulse-related behavior characterized by uncontrolled behavior; (2) Pathological impulsion or compulsive behavior associated with uncontrollable repetitive ideas; (3) uncontrollable forcing behavior related to forcing ideas; (4) Pathological behavior characterized by uncontrollable twitches, urge and compulsions; (5) pathological impulse-related behavior characterized by uncontrolled control; (6) Pathological impulsive and uncontrollable behaviors characterized by problems with cognitive control and cognitive flexibility; (7) Pathological impulses, runaway and compulsive behaviors associated with compulsive ideas and urge; and (8) pathological behavior characterized by uncontrolled and/or impaired motivation. In some embodiments, the impulse-related or uncontrollable behavior is selected from the group consisting of eating runaway, binge eating, emetic, craving, binge, self-injury, aggression, substance abuse, compulsive cleaning/cleansing, compulsive checkup, gambling addiction, dehairing, and skin scratch.

Each of the aspects and embodiments described herein can be used together unless expressly or clearly excluded from the context of the embodiments or aspects.

Drawings

The features of the present disclosure are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present disclosure will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the disclosure are utilized, and the accompanying drawings of which:

FIGS. 1A-1E show a decrease in the vmPFC-NAc linkage index in obese status (potentially severe clinical complications) of subjects with binge eating.

Fig. 1A is a graph showing the connection index between vmPFC and NAc (kruscarl-vorrichardson test=9.2052, p=.01). Post-pairwise comparisons showed a decrease in the left-right connectivity index for obesity (n=13) compared to the thin (n=19) and WNL (n=16) queues.

Fig. 1B is a diagram showing the connection index between vmPFC and NAc (kruscarl-voricor test=14.375, p=.0007). Post-pairwise comparisons showed a decrease in the left-right connectivity index for obesity (n=13) compared to the thin (n=19) and WNL (n=16) queues.

FIG. 1C shows a decrease in the probability of vmPFC-NAc streamlines in obese subjects. The group average plot of vmPFC-NAc streamline probability (yellow for high; red for low) is shown in sagittal, coronal and axial views. Probabilistic fiber bundle imaging results for individual subjects were transformed into standard MNI152 c space and averaged in the cohort. The destination area is defined in the standard space. An NAc mask (white) adapted from the CIT168 subcortical in vivo probability map and a vmPFC mask (green) defined based on the Harvard-Oxford cortical map.

Fig. 1D is a graph showing that BMI is inversely related to the right (ρ= -, 51, p <.01, double tail) and left (ρ= -, 48, p <.01, double tail) vmPFC-NAc connection indices throughout the bingo trend queue (n=37).

FIG. 1E is a graph showing that in the WNL queue, BMI is not related to either left (blue) or right (red) vmPFC-NAc connectivity index. N.s. =insignificant. * =p <.05.* = P <.01.* P <.001.

Figures 2A-2D show that vmPFC thickness is inversely related to BMI in subjects with binge eating.

Fig. 2A is a graph showing that the obese queue exhibited significantly lower right vmPFC thickness (p=.02) than the lean queue.

Fig. 2B is a graph showing that the difference in left vmPFC thickness between obese and lean queues did not reach statistical significance (p=.15). Throughout the binge eating trend cohort (n=37).

Fig. 2C is a graph showing that BMI was found to be inversely related to right-side vmPFC thickness (ρ= -, 42, p <.05, double tail) and edge inversely related to left-side vmPFC thickness (ρ= -, 29, p=.09, double tail).

Fig. 2D is a graph showing that the vmPFC-NAc connection index is correlated with the right-side vmPFC thickness (ρ=.50, p <.01, double tail) and the left-side edge correlation (ρ=.31, p=.06, double tail). N.s. =insignificant. * =p <.05.* = P <.01.

Figures 3A-3D show that depression scores in subjects with binge eating do not explain the correlation of vmPFC-NAc linkage index and vmPFC thickness with BMI.

Fig. 3A is a graph showing the increase in Beck's Depression Inventory, BDI score (p=.02) of the Beck depression self-rating scale of the obese cohort compared to the lean cohort.

Fig. 3B is a graph showing the presence of depression (i.e., bdi+.10) significantly different between obese and lean queues (χ2=6.3; p=0.01).

Fig. 3C is a graph (connection index |bdi) showing that correlation between BMI and the left (ρ= -, 44, p=.008, double tail) and right (ρ= -, 52, p=.002, double tail) vmPFC-NAc connection indices remains significant after including BDI score as a covariate.

Fig. 3D is a graph (vmPFC thickness |bdi) showing that the right vmPFC thickness also remains significant (ρ= -, 42, p=.01, double tailed) after including the BDI score as a covariate. N.s. =insignificant. * =p <.05.* = P <.01.

Figures 4A-4J show that depression scores in subjects with binge eating do not explain the correlation of vmPFC-NAc linkage index and vmPFC thickness with BMI.

Fig. 4A shows the group mean streamline probabilities from lower (red) to higher (yellow) in the canonical dataset of 178 HCP subjects between NAc and vmPFC.

Fig. 4B shows a group mean profile of the NAc shell (red) and core (blue) subregions in standard space based on the distribution of these streamlines in 178 subjects from the human linked group plan. Voxels present in less than 40% of subjects were excluded.

Fig. 4C is a graph showing that in subjects with binge eating (n=37), significantly higher normalized streamline counts were also observed between vmPFC and left (P <.001) and right (P <.001) NAc shells compared to the NAc core.

Fig. 4D is a graph showing that vmPFC-NAc shell rsFC is significantly lower (p=.04) in the obese queue.

Fig. 4E is a diagram showing a negative correlation (ρ= -.36, p=.04) with BMI.

Fig. 4F is a case display based on target definition of a subject-specific NAc loop using high resolution clinical diffusion MRI data from obese subjects undergoing reactive DBS as part of a first human clinical trial. The two most distal electrode contacts of subject 1 cover the NAc subregion (red, NAc shell) with vmPFC interaction as defined by more robust fiber bundle imaging on one side.

Fig. 4G is a case display based on target definition of a subject-specific NAc loop using high resolution clinical diffusion MRI data from obese subjects undergoing reactive DBS as part of a first human clinical trial. The two most distal electrode contacts of subject 2 cover the NAc subregion (red, NAc shell) with the vmPFC interaction defined by the more robust fiber bundle imaging on both sides.

Fig. 4H shows that only the left electrode contact is used to deliver the initial active stimulus. These electrode contacts cover the left NAc subregion of subject 2 (but not subject 1).

Fig. 4I is a graph showing that there was no significant change in the average elos-18 frequency score for subject 1 after the initiation of left stimulation (as shown in fig. 4H).

Fig. 4J is a graph showing that after initiation of left-hand active stimulation (as shown in fig. 4H), subject 2 had a significantly reduced average elos-18 frequency score (59.4% reduction relative to baseline; u=265, p=.001). N.s. =insignificant. * =p <.05.* = P <.01.* P <.001.

Detailed Description

The following description and examples detail embodiments of the present disclosure.

It is to be understood that this disclosure is not limited to the particular embodiments described herein, and as such, may vary. Those skilled in the art will recognize that there are variations and modifications of the present disclosure that are encompassed within its scope.

All terms are intended to be interpreted as they will be understood by those skilled in the art. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.

The section headings used herein are for organizational purposes only and are not to be construed as limiting the subject matter described.

Although various features of the disclosure may be described in the context of a single embodiment, the features can also be provided separately or in any suitable combination. Conversely, although the present disclosure may be described herein in the context of separate embodiments for clarity, the present disclosure may also be implemented in a single embodiment.

Definition of the definition

The following definitions supplement the definitions in the art and relate to the present application and are not to be construed as being any relevant or irrelevant cases, such as any commonly owned patent or application. Although any methods and materials similar or equivalent to those described herein can be used in the practice of the test in the present disclosure, the preferred materials and methods are described herein. Thus, the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting.

In this application, the use of the singular includes the plural unless specifically stated otherwise. It must be noted that, as used in the specification, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise.

In this application, the use of "or" means "and/or" unless stated otherwise. The terms "and/or" and "any combination thereof" and grammatical equivalents thereof as used herein may be used interchangeably. These terms may indicate that any combination is specifically contemplated. For illustrative purposes only, the following phrases "A, B and/or C" or "A, B, C, or any combination thereof," may mean "a alone; b alone; c alone; a and B; b and C; a and C; and A, B and C). The term "or" may be used in combination or separately unless the context specifically refers to separate use.

Furthermore, the use of the terms "include" and other forms, such as "comprises," "comprising," and "including," are not limiting.

Reference in the specification to "some embodiments," "an embodiment," "one embodiment," or "other embodiments" means that a particular feature, structure, or characteristic described in connection with the embodiments is included in at least some embodiments, but not necessarily all embodiments, of the disclosure.

As used in this specification and the claims, the words "comprise" (and any form of comprising, such as "comprises") and "comprising" (and any form of having, such as "have" and "have"), include (and any form of comprising, such as "include" and "include") or "contain (and any form of containing, such as" contain "and" contain ") are inclusive or open ended, and do not exclude additional, unrecited elements or method steps. It is contemplated that any of the embodiments discussed in this specification may be implemented with respect to any method or composition of the present disclosure, and vice versa. Furthermore, the compositions of the present disclosure may be used to implement the methods of the present disclosure.

SUMMARY

Obesity is an increasing global public health crisis caused, at least in part, by uncontrolled feeding of readily available, highly palatable and refined foods (Stice et al, 2013). The loss of inhibitory control of how much a person eats has been repeatedly correlated with the intractability of obesity, even with the most aggressive treatment available (Lavagnino et al, 2016a; white et al, 2010). Binge eating may be the most extreme branch of behavior that loses inhibitory control in view of the amount of food consumed in one episode (American Psychiatric Association,2013; ho et al, 2015). Perhaps not unexpectedly, obesity occurs more commonly 3-6 times in binge-fed subjects (Kessler et al, 2013). Binge eating not only worsens the outcome of treatment for obesity, but also predisposes individuals to impaired metabolic function, psychotic complications and impaired quality of life, in part due to the high rate of type 2 diabetes and the debilitation of food by excessive concern (Chao et al 2016; grucza et al 2007; mcCuen-Wurst et al 2018). Thus, examining the neural basis of obesity and binge eating may elucidate shared neural circuits, reveal previously unknown disease mechanisms, and guide neuromodulation interventions tailored for these common co-existence conditions (Stice et al, 2013).

Deep Brain Stimulation (DBS) of nucleus accumbens (NAc) is a neuromodulation strategy that has shown promise in improving binge eating behavior and inducing weight loss in mice (Halpern et al, 2013; wu et al, 2018). Single cases have also been reported in human subjects suffering from obesity that undergo NAc DBS and undergo restorative inhibitory control and significant weight loss (management et al, 2010; tronnier et al, 2018). These effects may involve retrograde activation of the ventral inner prefrontal cortex (vmPFC) since during NAc DBS, an induced increase in c-Fos expression and an increased blood oxygenation level dependent signal are observed, respectively, in vmPFC, especially in small and large animal species (Cho et al, 2019; vassoler et al, 2013). These findings are not surprising, as the vmPFC-NAc loop has been repeatedly described as mediating, inter alia, inhibitory control of food (Richard and Berridge,2013; vassoler et al, 2013). Pharmacological inactivation of rat vmPFC results in behavioral disinhibition in the suggested appetite task (Ghazizadeh et al 2012). In contrast, topical vmPFC activation suppresses appetite feeding (Richard and Berridge, 2013). Thus, the therapeutic effect of NAc DBS in the co-treatment of obesity and binge eating may be due, at least in part, to modulation of vmPFC and NAc interactions (Bossert et al, 2012; cartmell et al, 2019; pierce and Vassoler, 2013).

Methods of targeting neural circuits

Multimode imaging analysis is provided herein to examine and target the putative loop. Provided herein are (i) investigation of the relationship between Body Mass Index (BMI), vmPFC-NAc structural attachment and vmPFC thickness in women undergoing binge eating, taking into account available clinical and behavioral covariates; (ii) Using the most advanced canonical diffusion MRI dataset from the human linked group project (HCP) to locate the convergence position of streamlines between NAc and vmPFC within NAc; (iii) Applying the localization to a clinical sample of a female undergoing binge eating; (iv) Testing the resting state functional linkage (rsFC) of the uncovered NAc target region and its relationship to obesity in the same binge eating sample to examine the functional relevance of the NAc subregion; and (v) assessing the feasibility of directly modulating the vmPFC-NAc loop by probabilistic fiber bundle imaging guidance in two subjects experiencing NAc-reactive DBS as part of a first human clinical trial for loss of diet in refractory obesity (clinical trimals. Gov identification number: NCT 03868670) (Wu et al 2020).

The first study is provided herein to (i) discover a condition-specific effect on the vmPFC-NAc loop associated with obesity in subjects with binge eating, (ii) define a loop-based NAc target for obesity based on a specific interaction maintained with vmPFC, and (iii) demonstrate the feasibility of applying the method directly to loop-based targeting.

The study gave timely initial human trials of reactive NAc DBS in patients with obesity and with eating runaway behavior (e.g., binge eating) (ClinicalTrials gov identification number: NCT 03868670) (Wu et al 2020). Functional neuroimaging studies performed in adult volunteers over ten years have revealed a correlation between forehead She Huodong (including vmPFC) and BMI (Geha et al, 2017; volkow et al, 2009). Nevertheless, there are few reports of the association between vmPFC loops and BMI in structural imaging studies. Extensive white matter microstructural changes have been reported in subjects with binge eating (He et al, 2016); however, the white matter changes characteristic of individuals suffering from binge eating and obesity have not been explored.

The present disclosure provides evidence for the first time that supports the following: the structural connection between NAc and vmPFC was reduced in women with obesity and binge eating compared to thin women with binge eating only and WNL control. This finding may indicate that vmPFC-NAc is disturbed in subjects with obesity and binge eating. Furthermore, the finding that BMI is inversely related to vmPFC-NAc fiber bundle imaging-CI supports the damage, and possibly even vulnerability, of the brain circuit in obese queues (Ghazizadeh et al 2012; rapuano et al 2020). The present disclosure shows that cortical thickness in the vmPFC region in the obese cohort decreases and is inversely related to BMI. It may not be surprising that in obese queues the vmPFC-NAc loop perturbation is accompanied by thinning of vmPFC, which may be representative of the obesity liability of some subjects. Evidence of this correlation in terms of correlation between vmPFC-NAc fiber bundle imaging-CI and vmPFC thickness is provided herein.

Clinical measures of depression, anxiety and mood regulation are reported to be impaired in obese subjects with a history of binge eating (Grucza et al, 2007; kessler et al, 2013). When the obese cohort was compared to the lean cohort using the psychosis scale (psychiatric battery of scale), the only variable that differed between these cohorts was found to be the BDI score. Furthermore, the vmPFC-NAc loop is repeatedly involved in mediating an emotional state (Riva-Posse et al, 2018; vassoler et al, 2013; winnedoff et al, 2013). However, this measurement of depression does not fully explain the critical finding that BMI negatively affects the structural connection of the vmPFC-NAc circuit.

Using diffusion MRI data from HCPs and single case studies previously, it was found that based on whole brain connectivity spectra, probabilistic fiber bundle imaging can be used to subdivide human NAc into core (dorsal side) and shell (ventral side) sub-regions (cartsell et al 2019). This method has also been successfully applied to other brain regions (Kakusa et al 2021). In preclinical models, changes in neuronal activation in vmPFC following NAc DBS have been demonstrated across species (Cho et al, 2019; vassoler et al, 2013).

The disclosure provided herein demonstrates that two NAc subregions can be distinguished based on their connection to separate vmpfcs using probabilistic fiber bundle imaging; wherein the NAc shell concentrates most of the vmPFC streamlines in subjects with binge eating, possibly representing a functionally related NAc subregion with obesity. In both HCP subjects and binge eating trend cohorts, the sub-region defined by this fiber bundle imaging was found to be the region with higher numbers of vmPFC streamlines. This is consistent with the findings of previous use of synapse-tracing and CLARITY-based 3D histological fiber bundle imaging in mice and diffusion MRI fiber bundle imaging in humans, as it has also been demonstrated in rodents and primates (bosert et al 2012; cartsell et al 2019; haber et al 2006).

The functional correlation of the NAc shell subregion with a greater number of vmPFC streamlines is supported by the rsFC reduction of the vmPFC-NAc shell loop in the obese queue. Furthermore, the negative correlation of BMI and rsFC in this circuit further supports disease-specific changes. This is consistent with empirical imaging assays and computational models previously found to be robust in relation between fiber bundle imaging and rsFC in the presence of structural connections (honeyy et al 2009).

The disclosure provided herein demonstrates the feasibility of immediate clinical application of this novel targeting approach using clinical diffusion MRI. The feasibility of targeting the NAc shell was confirmed by evaluating lead placement incorporated into post-operative imaging of loop-based targets. Indeed, modulating the circuit with reactive DBS shows a not surprising causal relationship between the circuit and the subject's inhibitory control of feeding, since there is a preliminary recovery of self-control. This reproducibility of target definition based on the circuit using different clinical diffusion MRI parameters (including widely available diffusion MRI protocols) will pave the way for the wide application of this targeting approach to modulate vmPFC-NAc circuits. The ultimate impact of these findings is to achieve novel, personalized, loop-based targeting to optimize neuromodulation therapy for subjects with refractory obesity and binge eating.

Every maximum numerical limitation given throughout this specification includes every lower numerical limitation, as if such lower numerical limitations were expressly written herein. Every minimum numerical limitation given throughout this specification will include every higher numerical limitation, as if such higher numerical limitations were expressly written herein. Every numerical range given throughout this specification will include every narrower numerical range that falls within such broader numerical range, as if such narrower numerical ranges were all expressly written herein.

All patent applications, websites, other publications, accession numbers, etc. cited above or below are incorporated by reference in their entirety for all purposes to the same extent as if each individual item was specifically and individually indicated to be incorporated by reference. If different versions of the sequence are associated with accession numbers at different times, then the version associated with accession numbers at the date of the effective submission of the present application is intended. Valid commit date means the actual commit date or the earlier of the commit dates of the priority applications (if available, the accession numbers). Also, if different versions of a publication, web site, etc. are released at different times, then that release is the most recent release on the effective filing date of the application is intended unless indicated otherwise. Any feature, step, element, embodiment, or aspect of the disclosure may be used in combination with any other item unless specifically indicated otherwise.

Although the present disclosure has been described in some detail by way of illustration and example for purposes of clarity and understanding, it will be apparent that certain changes and modifications may be practiced within the scope of the appended claims.

Examples

These examples are provided for illustrative purposes only and do not limit the scope of the claims provided herein.

In the following examples, multimode imaging was used to study the vmPFC-NAc loop under disease conditions, and reduced vmPFC-NAc structural junctions and vmPFC thickness were found. Furthermore, vmPFC-NAc interactions are more prominent in the NAc subregions of obese subjects with reduced vmPFC functional linkage. The two binge eating prone obese subjects were invasively regulated for NAc. Subjects with stimuli covering sub-areas with more prominent vmPFC interactions reported significant improvements in food-specific inhibition control. Rescue of such circuit abnormalities as disclosed herein may restore control over food selection in obesity and related disorders.

EXAMPLE 1 MRI data and Pre-treatment

The MRI acquisition parameters are summarized in table 1.

Table 1. Summary of mri acquisition parameters

* For subject 2, images were acquired in a large caliber clinical scanner.

Including MRI data from three different queues: (1) Diffusion, structural and functional resting state MRI data on a 3T MRI scanner (Discovery MR750, GE Healthcare, milwauki, wisconsin) from 61 women recruited by the stent diet disorder program (Stanford Eating Disorders Program); (2) Obtaining a canonical diffusion MRI dataset from 178 unrelated subjects HCPs, who have been ultra-high resolution acquired on a "Magnetom"7T MRI scanner (Siemens Medical Systems, erlang, germany), from a publicly available S1200 WashU-Minn-Ox HCP dataset (glass et al, 2013; sotoropoulos et al, 2013; van Essen et al, 2012); (3) Diffusion and structural clinical 3T MRI data from two subjects with electrode placement in NAc (for more details see clinical application assessment and case description). To evaluate reproducibility of results using different parameters, two subjects from a reactive DBS clinical trial for loss of diet in refractory obesity were subjected to a high resolution diffusion MRI protocol. In view of the size of the subject, a diffusion sequence requires a scan time of 30 minutes using a large caliber MRI scanner.

Preprocessing of T1 weighted images involves the use of advanced normalization tools (Advanced Normalization Tool, ANT) cortical thickness tubing (pipeline) (Avants et al 2009). Through this conduit, each image is run using the existing template image (figshare. Com/armics/ants_antsr_brain_templates/915436) and the associated anatomical priors to generate a cortical thickness map in subject space. The pipeline involves a number of processing steps including bias correction (Tustinon et al, 2010), brain extraction, n-tissue segmentation and spatial normalization (Avants et al, 20111 a; 20110 b). Although diffusion MRI data from clinical cohorts and clinical trial subjects were pre-processed to prepare images for probabilistic fiber bundle imaging using the FSL suite (Andersson et al, 2003; smith et al, 2004), normative HCP diffusion MRI data has been pre-processed (using a minimal pre-processing pipeline). The motion and geometric distortion of the diffusion weighted image is corrected using "topup" and "eddy" functions (similar to those applied in the HCP's preprocessing pipeline). For each subject, the diffusion and T1 weighted images were co-registered using boundary-based registration. Resting fMRI scans from binge eating trend cohorts were pre-processed using fMRI prep 1.2.3 (Esteban et al, 2019). Briefly, preprocessing of functional images involves skull dissection, co-registration with T1 reference images, and head motion and susceptibility distortion correction. After removing the unsteady volumes and spatial smoothing with a 6mm FWHM isotropic gaussian kernel, the ICA-AROMA was used to identify motion-related noise components in the OLD signal (Pruim et al 2015). The root mean square variance (DVARS) of voxels of the frame-by-Frame Displacement (FD) and time derivative of the time history is calculated (Power et al 2012; 2014). Global signals were extracted within cerebrospinal fluid (CSF), white Matter (WM), gray Matter (GM) and whole brain mask. The pre-processed BOLD output from fMRIPrep is denoised using XCP Engine 1.0 using estimated confounding parameters (Ciric et al, 2017; parkes et al, 2018). This includes zero-averaging and removing any linear or quadratic trend, and temporal filtering using a first order butterworth band-pass filter (0.01-0.08 Hz). Then, after these preliminary preprocessing steps, a hybrid regression is performed on the ICA-AROMA noise component and the average WM, CSF and global signal regression. All regressions are bandpass filtered to preserve the same frequency range as the data to avoid frequency dependent mismatches (Ciric et al, 2017).

Example 2 demographic data, clinical data, and behavioral data for binge eating cohorts

Subject consent was obtained according to the declaration of helsinki (Declaration of Helsinki) and approved by the institutional ethical committee (institutional ethical committee) (IRB-35204). Available clinical and behavioral data from 61 women (average age=26±5.3 years; bmi=26.8±7.8) were analyzed, 37 of which had binge eating, defined as eating a large amount of food in a short period of time at least once a week with eating a loss of dietary sensation in the preceding 6 months (i.e., binge eating trend queue; average age=26±5.6 years; bmi=27.9±8.5; binge eating frequency=2.7±1.4 episodes/week) (American Psychiatric Association, 2013). The number of binge eating episodes per week was assessed using eating disorder checks, a standardized diagnostic interview (Fairburn and Cooper, 1993). Beck depression self-assessment (BDI) and Beck anxiety questionnaires were used to screen for depression and anxiety, respectively (Beck et al 1961,1988). Mood disorders were assessed using a mood disorder scale (Difficulties in Emotion Regulation Scale) (Gratz and Roemer, 2004). Subjects were initially divided into the following three cohorts (table 2): (1) Within Normal Limits (WNL) queue (n=16): BMI <25, average weekly binge eating episodes less than once in the previous 6 months; (2) lean binge eating trend cohort (n=19): BMI <25, at least once a week binge eating episode (called thin queue); (3) obesity binge eating trend queue (n=13): BMI >30, at least once a week of binge eating onset (known as obesity cohort). Clinical and behavioral variables that distinguish between lean and obese cohorts were further studied.

TABLE 2 summary of demographic, behavioral, and imaging measurements for the obese, lean, and within-normal (WNL) cohorts WNL cohort binge eating liabilities cohort

* Unless otherwise specified, the Mannheim U test is used.

The krueschel-wales rank sum test. />Checking, df=1. Post pair wise comparisons were made using the mann-whitney U test and the P values adjusted as appropriate for error discovery rate:

^a WNL queues-thin queues not equal to obese queues

EXAMPLE 3 probabilistic fiber bundle imaging

Probabilistic fiber bundle imaging was used to evaluate the connection between NAc and vmPFC. The NAc mask was defined according to a standard T1 MNI152 09c template adapted from CIT168 subcortical in vivo probability map (Pauli et al, 2018), whereas the vmPFC mask was defined using the Harvard-Oxford brain map as previously used (Dunlop et al, 2017). The registration is performed using an Advanced Normalization Tool (ANT), which consists of two consecutive steps of linear and non-linear registration between the brain of the subject and the MNI brain. In a third step, the MNI-defined ROI is registered to the space of the subject.

Bayesian estimation of diffusion parameters obtained using sampling techniques in FSL (BEDPOSTX) is used to sample the probability distribution of diffusion parameters at each voxel, taking into account at most three cross fiber directions within the voxel (Behrens et al, 2007). Fiber tracking was performed with Probtrackx2 of FSL using distance correction and each NAc voxel as seed and vmPFC as target (Jenkinson et al 2012). A total of 5000 seed points were used to generate streamlines from each seed voxel, and only streamlines reaching the target were retained for further analysis. The results of Probtrackx are summarized in a "streamline probability" graph, which gives the probability of each seed voxel reaching the target. waytotal represents the total number of streamlines from a given seed reaching the target.

EXAMPLE 4 measurement of cortical thickness

The same vmPFC mask was used in a cortical thickness analysis, which was performed to assess whether BMI was associated with vmPFC thinning in subjects with binge eating. Registration between the MNI template and the template used in cortical thickness tracts was performed using ANT (Avants et al 2009). These transformations consist of template-to-subject transformations from the pipeline to transform the vmPFC tag into the subject space. Notably, cortical thickness analysis was performed using public Scripts (https:// gitsub. Com/ANTsX/ANTs/tree/master/Scripts) from Penn image computing and science laboratories; while fiber bundle imaging-based analysis is performed using internally developed scripts, so a slightly modified different normalization method is applied; both use ANT. These subject spatial labels are then used to calculate the average cortical thickness of vmPFC using the voxel-by-voxel cortical thickness map.

Example 5 Fulvo nucleus segmentation

Fiber bundle imaging was used to subdivide each of 356 nacs (178 subjects) from the canonical HCP dataset based on each individual distribution of streamlines between nacs and vmPFC to define regions of denser localization of the vmPFC streamlines within the nacs.

This analysis was performed to define the NAc subregions in the canonical HCP data, which were then applied to the binge eating trend queue. The mean streamline probability map of NAc voxels to vmPFC was calculated to obtain a normalized weighted average group map of streamline probabilities between NAc and vmPFC in 178 HCP subjects. A connection matrix is also generated for each subject. The connection matrix stores the number of streamlines between rows (seeds) and columns (targets) and can be used for blind classification. Based on these connection matrices, each NAc is partitioned using k-means. This non-hypothesized method assigns each voxel to one of the two clusters using successive iterations without applying external spatial constraints. For the case of greater similarity between voxels in the streamline counts, the algorithm cannot identify two different clusters. The k-means was evaluated to identify the frequency of two different NAc clusters based on the connection of NAc to vmPFC. The resulting clusters for each subject were transformed into standard MNI space and concatenated into a reference atlas of nacs based on the linkage of nacs to vmPFC. This is useful for pre-operative consultation and helps to plan stereotactic surgery targeting NAc without the use of advanced diffusion MRI techniques. Finally, canonical clusters are co-registered with clinical cohort MRI images to evaluate the number of streamlines from each cluster (seed) to vmPFC.

Example 6 resting State functional connection analysis

Resting state functional link analysis was performed on pre-processed resting state fMRI data of binge eating trend queues using DPABI 4.3/DPARSF based on statistical parameter mapping (SPM 12, www.fil.ion.ucl.ac.uk/SPM) (Yan et al, 2016). Three subjects were excluded due to excessive motion (as measured by 1) average FD >0.2mm, 2) more than 20% FD exceeding 0.2mm, or 3) any FD >5 mm) (Parkes et al, 2018). A seed-based method was performed to examine rsFC in the binge eating trend queue (n=34) by calculating the same vmPFC and NAc mask as defined above and rsFC between each fiber bundle imaging defined NAc subregion. Functional link values for all subjects were extracted and used for further correlation analysis.

EXAMPLE 7 statistical analysis

The strength of the connection between the seed and the target is expressed as the fiber bundle imaging connection index (fiber bundle imaging-CI), as defined in previous studies by Tschentscher et al by the following formula: log (waytotal)/log (5000 x V seeds) (Tschentscher et al, 2019). The walltotal produced by fiber bundle imaging is logarithmically transformed and divided by the logarithm of the product of the sample flow line (5000) generated in each seed voxel and the number of voxels in the corresponding seed mask (vseeds). The logarithmic transformation increases the likelihood of reaching normal, which is tested using the Charpy-Wilker test (Royston, 1992). Statistical analysis was performed using RStudio version 1.2.5042 (RStudio, inc.). The krueschel-wales test was used to compare vmPFC-NAc fiber bundle imaging-CI between obese, lean and WNL queues. The man-whitney U test was used to compare vmPFC thickness and rsFC between NAc (and subregions) and vmPFC in obese and lean queues. The age-corrected spearman correlation coefficient (ρ) between BMI and (1) fiber bundle imaging-CI, (2) vmPFC thickness and (3) rsFC was calculated separately. Correction for age is applied taking into account the previously described effects of age on white matter tract and cortical thickness (Davis et al 2009; salat et al 2004). The mann-whitney U test was used to compare corrected streamline numbers between vmPFC and NAc subregions in the binge eating trend queue. Significance was defined by P <0.05 for all tests with error discovery rate (FDR) correction for multiple comparisons (where applicable).

Example 8 evaluation of clinical application and case Specification

After verifying participation of vmPFC-NAc fiber bundle imaging-CI and vmPFC thickness in binge eating prone subjects with obesity, and defining optimal targets within the NAc to modulate the vmPFC-NAc loop in high resolution dataset, the feasibility of invasive stimulation of the NAc subregion by fiber bundle imaging guidance was assessed. A high resolution diffusion preoperative MRI protocol was used in two morbid obese subjects with binge eating disorders who received bilateral NAc-reactive deep brain stimulation in a first human clinical trial (MRI protocol details are summarized in table 1). The frequency of uncontrolled diet related sensations or behaviors was examined for the previous 28 days using the diet runaway scale (ELOCS-18) (Blomquist et al, 2014). Baseline and ELOCS-18 frequency scores after stimulation start were analyzed. Subject 1 was a 45 year old female with an average eating disorder scale (elos-18) frequency score of 6.7 (bmi=46 kg/m 2). Subject 2 was a 56 year old female meeting the criteria for morbid obesity (bmi=48.9 kg/m 2) and having an average elos-18 frequency score of 13.3. Both subjects reported diverting food to deal with stress, however, they were not currently diagnosed with mood or anxiety disorders. BDI and positive and negative emotion scales (Positive and Negative Affect Schedule, PANAS) were used to screen for depression and evaluate changes in positive and negative emotion throughout the study (Beck et al, 1961; watson et al, 1988). According to the study protocol (ClinicalTrials. Gov identification number: NCT 03868670), these subjects underwent multiple treatment failures for obesity (including pharmacological and non-pharmacological therapies and Roux-en-Y gastric bypass) (Wu et al 2020). Informed consent was obtained prior to recruitment. The study protocol was approved by the institutional ethics committee (IRB-46563) and the U.S. FDA according to the study equipment exemption (IDE G180079).

The subject-specific fiber bundle imaging defined NAc subregions are delineated in the MRI of the subject and loaded into elents stereotactic planning software (bradlab, germany). The trajectory is planned according to published stereotactic coordinates for targeting NAc (Cartmell et al, 2019; kakusa et al, 2019; nauczyniel et al, 2013) and then optimized by direct targeting using fast gray acquisition T1 inversion recovery (fGATIR) and T1 images and fiber bundle imaging to ensure wire placement encompasses NAc subregions, with a high number of streamlines to vmPFC. Post-operative lead reconstruction is performed by co-registering the post-operative CT image with the pre-operative MRI to confirm lead placement associated with the NAc subregions defined by the personalized fiber bundle imaging. One-sided, under blind conditions, one-week reactive stimulation was initiated using only left-sided leads (consistent with preclinical work) (Halpern et al, 2013). Stimulation is triggered by a continuous biomarker detector (neurostimulator; model RNS-320; neuroPace) programmed to identify preliminary candidate biomarkers, which are defined offline as time-before-binge-eating capture, as previously described (Wu et al 2020). Stimulation parameters for each subject were defined based on responses to clinical acute stimulation assessments. Two subjects remained blind to the stimulus environment (e.g., stimulus versus sham stimulus). The mann-whitney U test was used to compare the ELOCS-18 frequency scores after baseline and onset of active stimulation. The baseline time point of the ELOCS-18 frequency score reflects the 28 day period in which the patient did not receive any stimulation. The stimulation time points of the ELOCS-18 frequency scores reflect a 28 day period covering the week of active stimulation.

Example 9 reduction of vmPFC-NAc structural attachment in obese queues

A summary of the image acquisition parameters is described in table 1. Available clinical and behavioral data from 61 women (average age=26±5.3 years; bmi=26.8±7.8) were analyzed, 37 of which had binge eating, defined as eating a large amount of food in a short period of time at least once a week with eating a loss of dietary sensation in the preceding 6 months (i.e., binge eating trend queue; average age=26±5.6 years; bmi=27.9±8.5; binge eating frequency=2.7±1.4 episodes/week) (American Psychiatric Association, 2013). Subjects were initially divided into the following three cohorts: (1) Within Normal Limits (WNL) queue (n=16): BMI <25, average weekly binge eating episodes less than once in the previous 6 months; (2) lean binge eating trend cohort (n=19): BMI <25, at least once a week binge eating episode (called thin queue); (3) obesity binge eating trend queue (n=13): BMI >30, at least once a week of binge eating onset (known as obesity cohort). The demographics, behavior, and imaging measurements of these queues are described in table 2.

The vmPFC-NAc fiber bundle imaging connection index (fiber bundle imaging-CI), as defined by Tschentscher et al (Tschentscher et al, 2019), was found to be significantly different in the queues in the right hemisphere (kruser-vorinostat test=9.2052, p=.01) and the left hemisphere (kruser-vorinostat test=14.375, p=.0007). Post pair comparison showed that obese cohorts had reduced right-side fiber bundle imaging-CI (u=182, p <.05; u=174, p <.01; fdr correction) and left-side (u=200, p <.01; u=187, p <.001; fdr correction) fiber bundle imaging-CI compared to lean cohorts and WNL cohorts. The correlation of behavior and imaging measurements with BMI in the WNL and binge eating trend queues is described in table 3.

TABLE 3 correlation of behavioral and imaging measurements with BMI in Within Normal Limit (WNL) and binge eating trend queues

* Spearman correction, correction for age.

The BMI was inversely related to right side vmPFC-NAc fiber bundle imaging-CI (ρ= -, 51, p <.01, double tail) and left side vmPFC-NAc fiber bundle imaging-CI (ρ= -, 48, p <.01, double tail) throughout a binge eating trend cohort, which included all subjects (n=37) that had a large number of food episodes taken at least once a week for a short period of time with a perceived loss of diet over the previous 6 months (fig. 1D). In the WNL queue, BMI was uncorrelated with right side vmPFC-NAc fiber bundle imaging-CI (ρ= 18, P= 51, double tail) or left side vmPFC-NAc fiber bundle imaging-CI (ρ= -, 21, P= 45, double tail) (FIG. 1E).

Example 10 vmPFC thickness reduction in obese queue

Based on structural connection findings in the binge eating trend queue showing negative correlation of BMI with vmPFC-NAc fiber bundle imaging-CI, it was evaluated whether there was a relationship between vmPFC thickness and BMI. Indeed, the right obese cohort had a significantly reduced vmPFC thickness (u=183, p=.02) compared to the lean cohort (fig. 2A), but the left reduced vmPFC thickness did not reach statistical significance (u=162, p=.15) (fig. 2B). In the binge eating trend queue, BMI is related to right vmPFC thickness (ρ= -, 42, p <.05, double tail) and edge-related to left vmPFC thickness (ρ= -, 29, p=.09, double tail) (fig. 2C). Finally, vmPFC-NAc fiber bundle imaging-CI is significantly related to right side vmPFC thickness (ρ=.50, p <.01, double tail) and edge-related to left side vmPFC thickness (ρ=.31, p=.06, double tail) (fig. 2D).

Example 11 depression score fails to account for correlation of vmPFC-NAc with BMI

Depression scores (becker depression self-rating scale, BDI) increased significantly in the obese cohort compared to the lean cohort and WNL cohort (u=62, p=.02) (fig. 3A). There were no differences in the measurement of binge eating frequency, anxiety and mood adjustment problems between lean and obese cohorts (table 2). Not surprisingly, the presence of a classification of depression (measured here by bdi+.10) has a significant difference between the queues (χ2=6.3; p=.01); wherein the prevalence of depression was higher in the obese cohort compared to the lean cohort (FIG. 3B; table 2). Nevertheless, after including depression as a covariate, the correlation between BMI and left side vmPFC-NAc fiber bundle imaging-CI (ρ= -44, p=.008, double tail) and right side vmPFC-NAc fiber bundle imaging-CI (ρ= -, 52, p=.002, double tail) (fig. 3C) and right side vmPFC thickness (fig. 3D) (ρ= -, 42, p=.01, double tail) was still significant. After controlling the effect of the depressive variable (partiaging out), the BMI no longer has an edge correlation with the left vmPFC thickness (ρ= -, 27, p=.11, two-tailed) (fig. 3D).

EXAMPLE 12 pooling of vmPFC-NAc streamlines within the NAc Shell subregion

After identification of vmPFC-NAc fiber bundle imaging-CI and reduced vmPFC thickness in the obese cohort, probabilistic fiber bundle imaging was performed on a canonical high resolution HCP 7T diffusion MRI dataset of 178 subjects to reveal more densely located regions of vmPFC streamlines within the NAc, or to reveal putative NAc subregions in the vmPFC-NAc loop that can be targeted (fig. 4A). In the canonical HCP dataset, the k-means successfully identified two subregions among all 356 nacs analyzed with different linkages to vmPFC (i.e., one per hemisphere for each subject in the canonical dataset) (fig. 4B). Most of the vmPFC streamlines converge in the intra-abdominal region within the NAc, similar to the hypothetical NAc shell (defined herein as the shell) described previously, while other subregions similar to the hypothetical NAc core (defined herein as the core) receive fewer vmPFC streamlines (Baliki et al, 2013; cartsell et al, 2019). This analysis was performed in the canonical HCP dataset to define the NAc subregion, which was then applied to the binge eating trend queue. The shell sub-regions as defined by probabilistic fiber bundle imaging in the high resolution canonical HCP dataset also exhibited higher normalized streamline counts to the left (u=363, p <.001) and right (u=367, p <.001) vmPFC compared to the NAc core in subjects with binge eating (when used as seed region in probabilistic fiber bundle imaging) (fig. 4C).

Example 13 reduced vmPFC-NAc Shell resting state functional ligation in obese queues

The hypothesis that this NAc target region (i.e., the NAc shell) exhibited reduced rsFC with vmPFC in the obese cohort was tested. Indeed, the vmPFC-NAc shell rsFC was found to be significantly lower in obese queues compared to lean queues (u=149, p=.04) (fig. 4D). Consistent with this assumption, vmPFC-NAc shell rsFC is also significantly correlated with BMI in the overall binge eating cohort (ρ= -, 36, p=.04) (fig. 4E).

Example 14 case description-direct targeting of vmPFC-NAc Loop in clinical Environment

To assess the feasibility of applying probabilistic fiber bundle imaging to target this NAc subregion (i.e., the NAc shell) where vmPFC streamlines are more densely located in a clinical setting, a pre-operative diffusion MRI protocol (summarized in table 1) was used. Two subjects with binge eating disorders who met the morbid obesity criteria received a high resolution diffusion MRI protocol (Wu et al 2020) before experiencing reactive DBS in a first human clinical trial (clinical signs: NCT 03868670). To confirm the position of the electrode implant, intraoperative and postoperative CT scans were used so that the electrode could be reconstructed and covered with loop-based targets derived from analysis using preoperative MRI. The NAc subregions (shells) in which the vmPFC streamlines are more robustly converging can be identified and targeted in two subjects (fig. 4F). The two most ventral contacts were successfully implanted into the single-sided (right) and double-sided NAc shells of subject 1 and subject 2 (fig. 4G).

Example 15 modulation of the vmPFC-NAc loop improves self-control of food selection

Reactive DBS was initiated in two subjects using only the left lead. Thus, it is apparent that the fiber bundle imaging defined vmPFC-NAc loop was regulated in subject 2, but not necessarily in subject 1. Subject 1 was started in monopolar mode using the second most ventral contact (3.0 mA; 80. Mu.s; 125Hz; charge density 3.0. Mu.C/cm 2). Subject 2 was started in bipolar mode using two ventral contacts (0.5 mA;80 μs;125Hz; charge density 0.5 μC/cm 2). The stimulation was delivered in two 5 second pulses triggered by a continuous detector (neurostimulator; model RNS-320; neuroPace) programmed to identify candidate biomarkers of binge eating (not including data), consistent with the method previously described (Wu et al, 2018). Both subjects remained blind to the stimulus and the sham stimulus. After initiation of active stimulation, the average eating control scale (elos-18) frequency score for subject 1 was reduced by 18% from baseline (u=192, p=.34), while the average elos-18 frequency score for subject 2 was reduced by 59.4% from baseline (u=265, p=.001). Subject 1 reported an effort to make better food selection after stimulation, but did not obtain any differences from baseline. In contrast, subject 2 reported that food selection was more controlled by a constant sensation within 24 hours of the onset of active stimulation. Specifically, the latter subjects reported the following: "I have not eaten peanut butter for two days. I have had no desire for it. May all be in my mind (laughing), but it is not desired. This is somewhat frightening, since I have previously consumed 3-4 times per day. "she later reported that only three meals per day had been consumed since the addition of the project, and that there was no prior urge to eat snacks between meals. Furthermore, no significant change in BDI or positive negative emotion scale (pannas) score was observed for either subject after initiation of active stimulation.

While the present disclosure has been particularly shown and described with reference to particular embodiments, some of which are preferred, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the present disclosure as disclosed herein.

Reference to the literature

American Psychiatric Association(2013).Diagnostic and Statistical Manual of Mental Disorders(American Psychiatric Association).

Andersson,J.L.R.,Skare,S.,and Ashburner,J.(2003).How to correct susceptibility distortions in spin-echo echo-planar images:application to diffusion tensor imaging.NeuroImage 20,870–888.

Avants,B.,Tustison,N.,and Song,G.(2009).Advanced Normalization Tools:V1.0.Insight J.681.

Avants,B.B.,Tustison,N.J.,Song,G.,Cook,P.A.,Klein,A.,and Gee,J.C.(2011a).A reproducible evaluation of ANTs similarity metric performance in brain image registration.NeuroImage 54,2033–2044.

Avants,B.B.,Tustison,N.J.,Wu,J.,Cook,P.A.,and Gee,J.C.(2011b).An Open Source Multivariate Framework for n-Tissue Segmentation with Evaluation on Public Data.Neuroinformatics 9,381–400.

Baliki,M.N.,Mansour,A.,Baria,A.T.,Huang,L.,Berger,S.E.,Fields,H.L.,and Apkarian,A.V.(2013).Parceling Human Accumbens into Putative Core and Shell Dissociates Encoding of Values for Reward and Pain.J.Neurosci.33,16383–16393.

Beck,A.T.,Ward,C.H.,Mendelson,M.,Mock,J.,and Erbaugh,J.(1961).An Inventory for Measuring Depression.Arch.Gen.Psychiatry 4,561–571.

Beck,A.T.,Epstein,N.,Brown,G.,and Steer,R.A.(1988).An inventory for measuring clinical anxiety:psychometric properties.J.Consult.Clin.Psychol.56,893–897.

Behrens,T.E.J.,Berg,H.J.,Jbabdi,S.,Rushworth,M.F.S.,and Woolrich,M.W.(2007).Probabilistic diffusion tractography with multiple fibre orientations:What can we gainNeuroImage 34,144–155.

Blomquist,K.K.,Roberto,C.A.,Barnes,R.D.,White,M.A.,Masheb,R.M.,and Grilo,C.M.(2014).Development and Validation of the Eating Loss of Control Scale.Psychol.Assess.26,77–89.

Bossert,J.M.,Stern,A.L.,Theberge,F.R.M.,Marchant,N.J.,Wang,H.-L.,Morales,M.,and Shaham,Y.(2012).Role of Projections from Ventral Medial Prefrontal Cortex to Nucleus Accumbens Shell in Context-Induced Reinstatement of Heroin Seeking.J.Neurosci.32,4982–4991.

Cartmell,S.C.,Tian,Q.,Thio,B.J.,Leuze,C.,Ye,L.,Williams,N.R.,Yang,G.,Ben-Dor,G.,Deisseroth,K.,Grill,W.M.,et al.(2019).Multimodal characterization of the human nucleus accumbens.NeuroImage 198,137–149.

Chao,A.M.,Wadden,T.A.,Faulconbridge,L.F.,Sarwer,D.B.,Webb,V.L.,Shaw,J.A.,Thomas,J.G.,Hopkins,C.M.,Bakizada,Z.M.,Alamuddin,N.,et al.(2016).Binge-eating disorder and the outcome of bariatric surgery in a prospective,observational study:Two-year results.Obesity 24,2327–2333.

Cho,S.,Hachmann,J.T.,Balzekas,I.,In,M.-H.,Andres-Beck,L.G.,Lee,K.H.,Min,H.-K.,and Jo,H.J.(2019).Resting-state functional connectivity modulates the BOLD activation induced by nucleus accumbens stimulation in the swine brain.Brain Behav.9,e01431.

Ciric,R.,Wolf,D.H.,Power,J.D.,Roalf,D.R.,Baum,G.L.,Ruparel,K.,Shinohara,R.T.,Elliott,M.A.,Eickhoff,S.B.,Davatzikos,C.,et al.(2017).Benchmarking of participant-level confound regression strategies for the control of motion artifact in studies of functional connectivity.NeuroImage 154,174–187.

Davis,S.W.,Dennis,N.A.,Buchler,N.G.,White,L.E.,Madden,D.J.,and Cabeza,R.(2009).Assessing the effects of age on long white matter tracts using diffusion tensor tractography.NeuroImage 46,530–541.

Donnelly,B.,Touyz,S.,Hay,P.,Burton,A.,Russell,J.,and Caterson,I.(2018).Neuroimaging in bulimia nervosa and binge eating disorder:a systematic review.J.Eat.Disord.6,3.

Dunlop,B.W.,Rajendra,J.K.,Craighead,W.E.,Kelley,M.E.,McGrath,C.L.,Choi,K.S.,Kinkead,B.,Nemeroff,C.B.,and Mayberg,H.S.(2017).Functional Connectivity of the Subcallosal Cingulate Cortex And Differential Outcomes to Treatment With Cognitive-Behavioral Therapy or Antidepressant Medication for Major Depressive Disorder.Am.J.Psychiatry 174,533–545.

Esteban,O.,Markiewicz,C.J.,Blair,R.W.,Moodie,C.A.,Isik,A.I.,Erramuzpe,A.,Kent,J.D.,Goncalves,M.,DuPre,E.,Snyder,M.,et al.(2019).fMRIPrep:a robust preprocessing pipeline for functional MRI.Nat.Methods 16,111–116.

Fairburn,C.G.,and Cooper,Z.(1993).The Eating Disorder Examination(12th edition).In Binge Eating:Nature,Assessment,and Treatment,(New York,NY,US:Guilford Press),pp.317–360.

Geha,P.,Cecchi,G.,Todd Constable,R.,Abdallah,C.,and Small,D.M.(2017).Reorganization of brain connectivity in obesity.Hum.Brain Mapp.38,1403–1420.

Ghazizadeh,A.,Ambroggi,F.,Odean,N.,and Fields,H.L.(2012).Prefrontal Cortex Mediates Extinction of Responding by Two Distinct Neural Mechanisms in Accumbens Shell.J.Neurosci.32,726–737.

Glasser,M.F.,Sotiropoulos,S.N.,Wilson,J.A.,Coalson,T.S.,Fischl,B.,Andersson,J.L.,Xu,J.,Jbabdi,S.,Webster,M.,Polimeni,J.R.,et al.(2013).The minimal preprocessing pipelines for the Human Connectome Project.NeuroImage 80,105–124.

Gratz,K.L.,and Roemer,L.(2004).Multidimensional Assessment ofEmotion Regulation and Dysregulation:Development,Factor Structure,and Initial Validation of the Difficulties in Emotion Regulation Scale.J.Psychopathol.Behav.Assess.26,41–54.

Grucza,R.A.,Przybeck,T.R.,and Cloninger,C.R.(2007).Prevalence and Correlates of Binge Eating Disorder in a Community Sample.Compr.Psychiatry 48,124–131.

Haber,S.N.,Kim,K.-S.,Mailly,P.,and Calzavara,R.(2006).Reward-Related Cortical Inputs Define a Large Striatal Region in Primates That Interface with Associative Cortical Connections,Providing a Substrate for Incentive-Based Learning.J.Neurosci.26,8368–8376.

Halpern,C.H.,Tekriwal,A.,Santollo,J.,Keating,J.G.,Wolf,J.A.,Daniels,D.,and Bale,T.L.(2013).Amelioration of binge eating by nucleus accumbens shell deep brain stimulation in mice involves D2 receptor modulation.J.Neurosci.Off.J.Soc.Neurosci.33,7122–7129.He,X.,Stefan,M.,Terranova,K.,Steinglass,J.,and Marsh,R.(2016).Altered White Matter Microstructure in Adolescents and Adults with Bulimia Nervosa.Neuropsychopharmacol.Off.Publ.Am.Coll.Neuropsychopharmacol.41,1841–1848.

Ho,A.L.,Sussman,E.S.,Zhang,M.,Pendharkar,A.V.,Azagury,D.E.,Bohon,C.,and Halpern,C.H.(2015).Deep Brain Stimulation for Obesity.Cureus 7,e259.

Honey,C.J.,Sporns,O.,Cammoun,L.,Gigandet,X.,Thiran,J.P.,Meuli,R.,and Hagmann,P.(2009).Predicting human resting-state functional connectivity from structural connectivity.Proc.Natl.Acad.Sci.106,2035–2040.

Jenkinson,M.,Beckmann,C.F.,Behrens,T.E.J.,Woolrich,M.W.,and Smith,S.M.(2012).FSL.NeuroImage 62,782–790.

Kakusa,B.,Saluja,S.,Tate,W.J.,Espil,F.M.,Halpern,C.H.,and Williams,N.R.(2019).Robust clinical benefit of multi-target deep brain stimulation for treatment of Gilles de la Tourette syndrome and its comorbidities.Brain Stimulat.12,816–818.

Kakusa,B.,Saluja,S.,Barbosa,D.A.N.,Cartmell,S.,Espil,F.M.,Williams,N.R.,McNab,J.A.,and Halpern,C.H.(2021).Evidence for the role of the dorsal ventral lateral posterior thalamic nucleus connectivity in deep brain stimulation for Gilles de la Tourette syndrome.J.Psychiatr.Res.132,60–64.

Kessler,R.C.,Berglund,P.A.,Chiu,W.T.,Deitz,A.C.,Hudson,J.I.,Shahly,V.,Aguilar-Gaxiola,S.,Alonso,J.,Angermeyer,M.C.,Benjet,C.,et al.(2013).The prevalence and correlates of binge eating disorder in the WHO World Mental Health Surveys.Biol.Psychiatry 73,904–914.

Lavagnino,L.,Arnone,D.,Cao,B.,Soares,J.C.,and Selvaraj,S.(2016a).Inhibitory control in obesity and binge eating disorder:A systematic review and meta-analysis of neurocognitive and neuroimaging studies.Neurosci.Biobehav.Rev.68,714–726.

Lavagnino,L.,Mwangi,B.,Bauer,I.E.,Cao,B.,Selvaraj,S.,Prossin,A.,and Soares,J.C.(2016b).Reduced Inhibitory Control Mediates the Relationship Between Cortical Thickness in the Right Superior Frontal Gyrus and Body Mass Index.Neuropsychopharmacology 41,2275–2282.

Mantione,M.,van de Brink,W.,Schuurman,P.R.,and Denys,D.(2010).Smoking Cessation and Weight Loss After Chronic Deep Brain Stimulation of the Nucleus Accumbens:Therapeutic and Research Implications:Case Report.Neurosurgery 66.

Marsh,R.,Stefan,M.,Bansal,R.,Hao,X.,Walsh,B.T.,and Peterson,B.S.(2015).Anatomical Characteristics of the Cerebral Surface in Bulimia Nervosa.Biol.Psychiatry 77,616–623.McCuen-Wurst,C.,Ruggieri,M.,and Allison,K.C.(2018).Disordered eating and obesity:associations between binge eating-disorder,night-eating syndrome,and weight-related co-morbidities.Ann.N.Y.Acad.Sci.1411,96–105.

Medic,N.,Ziauddeen,H.,Ersche,K.D.,Farooqi,I.S.,Bullmore,E.T.,Nathan,P.J.,Ronan,L.,and Fletcher,P.C.(2016).Increased body mass index is associated with specific regional alterations in brain structure.Int.J.Obes.2005 40,1177–1182.

Nauczyciel,C.,Robic,S.,Dondaine,T.,Verin,M.,Robert,G.,Drapier,D.,Naudet,F.,and Millet,B.(2013).The nucleus accumbens:a target for deep brain stimulation in resistant major depressive disorder.J.Mol.Psychiatry 1,17.

Parkes,L.,Fulcher,B.,Yücel,M.,and Fornito,A.(2018).An evaluation of the efficacy,reliability,and sensitivity of motion correction strategies for resting-state functional MRI.NeuroImage 171,415–436.

Pauli,W.M.,Nili,A.N.,and Tyszka,J.M.(2018).A high-resolution probabilistic in vivo atlas of human subcortical brain nuclei.Sci.Data 5,180063.

Pierce,R.C.,and Vassoler,F.M.(2013).Deep brain stimulation for the treatment of addiction:basic and clinical studies and potential mechanisms of action.Psychopharmacology(Berl.)229,487–491.

Power,J.D.,Barnes,K.A.,Snyder,A.Z.,Schlaggar,B.L.,and Petersen,S.E.(2012).Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion.NeuroImage 59,2142–2154.

Power,J.D.,Mitra,A.,Laumann,T.O.,Snyder,A.Z.,Schlaggar,B.L.,and Petersen,S.E.(2014).Methods to detect,characterize,and remove motion artifact in resting state fMRI.NeuroImage 84,320–341.

Pruim,R.H.R.,Mennes,M.,van Rooij,D.,Llera,A.,Buitelaar,J.K.,and Beckmann,C.F.(2015).ICA-AROMA:A robust ICA-based strategy for removing motion artifacts from fMRI data.NeuroImage 112,267–277.

Rapuano,K.M.,Laurent,J.S.,Hagler,D.J.,Hatton,S.N.,Thompson,W.K.,Jernigan,T.L.,Dale,A.M.,Casey,B.J.,and Watts,R.(2020).Nucleus accumbens cytoarchitecture predicts weight gain in children.Proc.Natl.Acad.Sci.117,26977–26984.

Richard,J.M.,and Berridge,Kent.C.(2013).Prefrontal cortex modulates desire and dread generated by nucleus accumbens glutamate disruption.Biol.Psychiatry 73,360–370.

Riva-Posse,P.,Choi,K.S.,Holtzheimer,P.E.,Crowell,A.L.,Garlow,S.J.,Rajendra,J.K.,McIntyre,C.C.,Gross,R.E.,and Mayberg,H.S.(2018).A connectomic approach for subcallosal cingulate deep brain stimulation surgery:prospective targeting in treatment-resistant depression.Mol.Psychiatry 23,843–849.

Royston,P.(1992).Approximating the Shapiro-Wilk W-test for non-normality.Stat.Comput.2,117–119.

Salat,D.H.,Buckner,R.L.,Snyder,A.Z.,Greve,D.N.,Desikan,R.S.R.,Busa,E.,Morris,J.C.,Dale,A.M.,and Fischl,B.(2004).Thinning of the Cerebral Cortex in Aging.Cereb.Cortex 14,721–730.

Smith,S.M.,Jenkinson,M.,Woolrich,M.W.,Beckmann,C.F.,Behrens,T.E.J.,Johansen-Berg,H.,Bannister,P.R.,De Luca,M.,Drobnjak,I.,Flitney,D.E.,et al.(2004).Advances in functional and structural MR image analysis and implementation as FSL.NeuroImage 23 Suppl 1,S208-219.

Sotiropoulos,S.N.,Jbabdi,S.,Xu,J.,Andersson,J.L.,Moeller,S.,Auerbach,E.J.,Glasser,M.F.,Hernandez,M.,Sapiro,G.,Jenkinson,M.,et al.(2013).Advances in diffusion MRI acquisition and processing in the Human Connectome Project.NeuroImage 80,125–143.

Stice,E.,Figlewicz,D.P.,Gosnell,B.A.,Levine,A.S.,and Pratt,W.E.(2013).The contribution of brain reward circuits to the obesity epidemic.Neurosci.Biobehav.Rev.37,2047–2058.

Tronnier,V.M.,Rasche,D.,Thorns,V.,Alvarez-Fischer,D.,Münte,T.F.,and Zurowski,B.(2018).Massive weight loss following deep brain stimulation of the nucleus accumbens in a depressed woman.Neurocase 24,49–53.

Tschentscher,N.,Ruisinger,A.,Blank,H.,Díaz,B.,and Kriegstein,K.von(2019).Reduced Structural Connectivity Between Left Auditory Thalamus and the Motion-Sensitive Planum Temporale in Developmental Dyslexia.J.Neurosci.39,1720–1732.

Tustison,N.J.,Avants,B.B.,Cook,P.A.,Zheng,Y.,Egan,A.,Yushkevich,P.A.,and Gee,J.C.(2010).N4ITK:Improved N3 Bias Correction.IEEE Trans.Med.Imaging 29,1310–1320.

Van Essen,D.C.,Ugurbil,K.,Auerbach,E.,Barch,D.,Behrens,T.E.J.,Bucholz,R.,Chang,A.,Chen,L.,Corbetta,M.,Curtiss,S.W.,et al.(2012).The Human Connectome Project:a data acquisition perspective.NeuroImage 62,2222–2231.

Vassoler,F.M.,White,S.L.,Hopkins,T.J.,Guercio,L.A.,Espallergues,J.,Berton,O.,Schmidt,H.D.,and Pierce,R.C.(2013).Deep Brain Stimulation of the Nucleus Accumbens Shell Attenuates Cocaine Reinstatement through Local and Antidromic Activation.J.Neurosci.33,14446–14454.

Volkow,N.D.,Wang,G.-J.,Telang,F.,Fowler,J.S.,Goldstein,R.Z.,Alia-Klein,N.,Logan,J.,Wong,C.,Thanos,P.K.,Ma,Y.,et al.(2009).Inverse Association Between BMI and Prefrontal Metabolic Activity in Healthy Adults.Obes.Silver Spring Md 17,60–65.

Watson,D.,Clark,L.A.,and Tellegen,A.(1988).Development and validation of brief measures of positive and negative affect:the PANAS scales.J.Pers.Soc.Psychol.54,1063–1070.

White,M.A.,Kalarchian,M.A.,Masheb,R.M.,Marcus,M.D.,and Grilo,C.M.(2010).Loss of control over eating predicts outcomes in bariatric surgery patients:a prospective,24-month follow-up study.J.Clin.Psychiatry 71,175–184.

Winecoff,A.,Clithero,J.A.,Carter,R.M.,Bergman,S.R.,Wang,L.,and Huettel,S.A.(2013).Ventromedial Prefrontal Cortex Encodes Emotional Value.J.Neurosci.33,11032–11039.

Wu,H.,Miller,K.J.,Blumenfeld,Z.,Williams,N.R.,Ravikumar,V.K.,Lee,K.E.,Kakusa,B.,Sacchet,M.D.,Wintermark,M.,Christoffel,D.J.,et al.(2018).Closing the loop on impulsivity via nucleus accumbens delta-band activity in mice and man.Proc.Natl.Acad.Sci.115,192–197.

Wu,H.,Adler,S.,Azagury,D.E.,Bohon,C.,Safer,D.L.,Barbosa,D.A.N.,Bhati,M.T.,Williams,N.R.,Dunn,L.B.,Tass,P.A.,et al.(2020).Brain-Responsive Neurostimulation for Loss of Control Eating:Early Feasibility Study.Neurosurgery.

Yan,C.-G.,Wang,X.-D.,Zuo,X.-N.,and Zang,Y.-F.(2016).DPABI:Data Processing&Analysis for(Resting-State)Brain Imaging.Neuroinformatics 14,339–351.

Claims

1. A method of targeting a neural circuit of a subject having impulse-related or uncontrollable behavior, the method comprising

(a) Generating a circuit-specific brain connection spectrum using a pattern of water diffusivity based on the MRI of the subject;

(b) Generating a loop-specific voxel-by-voxel connection matrix with the loop-specific brain connection spectrum to define a region of interest including a neural loop to be targeted;

(c) Segmenting the loop-specific voxel-by-voxel connection matrix using an algorithm to generate 3D subregions to identify neural loops to be targeted; and

(d) Targeting the identified neural circuit with neuromodulation therapy.

2. A method of treating a subject having impulse-related or uncontrollable behavior, the method comprising

(d) Treating the subject with neuromodulation therapy by targeting the identified neural circuit.

3. A method of personalizing neuromodulation therapy for a subject having impulse-related or uncontrollable behavior, the method comprising

(d) The neuromodulation therapy is personalized based on patient-specific neuromorphic imaging, brain connectivity, cortical thickness, clinical and behavioral complaints, previous responses to neuromodulation, and other available clinical and behavioral covariates of the subject.

4. A method according to any one of claims 1-3, wherein the MRI is diffusion, structural and/or functional resting state MRI.

5. A method according to any of claims 1-3, wherein the algorithm is selected from the group consisting of k-means clustering algorithms, independent component analysis, principal component analysis, random relaxation with decoder perturbation, inter-centroid random search and hierarchical clustering algorithms.

6. A method according to any one of claims 1-3, wherein the region of interest is selected from the group consisting of nucleus accumbens (NAc), inner capsular forelimbs (ALIC), subthalamic nuclei (STN), thalamus, ventral medial prefrontal cortex (vmPFC), anterior cingulate retrocortex (ACC), orbitofum (OFC)/frontal cortex (FP) and dorsal medial/dorsal lateral prefrontal cortex.

7. The method of claim 6, wherein the region of interest is a nucleus accumbens (NAc).

8. The method of any one of claims 1-3, wherein the identified neural circuit is selected from the group consisting of a ventral prefrontal cortex (vmPFC) -NAc circuit, a orbitofrontal/frontal polar cortex-ALIC circuit, an orbitofrontal/frontal polar cortex-thalamic circuit, a vmPFC-NAc circuit, an ACC-NAc circuit, a dorsal prefrontal cortex-vmPFC-NAc circuit.

9. The method of claim 8, wherein the identified neural circuit is a vmPFC-NAc circuit.

10. The method of any one of claims 1-3, wherein the neuromodulation therapy is invasive or non-invasive.

11. The method of claim 10, wherein the neuromodulation therapy is selected from the group consisting of deep brain stimulation, ultrasound focusing, transcranial magnetic stimulation, transcranial electrical stimulation, radio modulation, and nerve ablation.

12. A method according to any one of claims 1-3, wherein the impulse-related or uncontrollable behavior is selected from (1) pathological impulse-related behavior characterized by uncontrolled behavior; (2) Pathological impulsion or compulsive behavior associated with uncontrollable repetitive ideas; (3) uncontrollable forcing behavior related to forcing ideas; (4) Pathological behavior characterized by uncontrollable twitches, urge and compulsions; (5) pathological impulse-related behavior characterized by uncontrolled control; (6) Pathological impulsive and uncontrollable behaviors characterized by problems with cognitive control and cognitive flexibility; (7) Pathological impulses, runaway and compulsive behaviors associated with compulsive ideas and urge; and (8) pathological behavior characterized by uncontrolled and/or impaired motivation.

13. A method according to any one of claims 1-3, wherein the impulse-related or uncontrollable behavior is selected from the group consisting of eating runaway, binge eating, emetic, craving, binge, self-injury, aggressiveness, substance abuse, compulsive cleaning/cleansing, compulsive check, gambling addiction, dehairing and skin shaving.